The field of wireless communications is witnessing a significant shift towards the integration of Large Language Models (LLMs) to enhance reasoning capabilities and improve performance in various tasks. This direction is driven by the need for more intelligent and adaptive wireless systems that can operate effectively in dynamic and complex environments. The use of LLMs is enabling the development of modular, extensible, and interpretable architectures that can selectively query and interpret expert outputs, leading to improved classification and decision-making capabilities. Notably, LLMs are being applied to tasks such as automatic modulation classification, covert communications, and context-aware Wi-Fi roaming, achieving significant improvements over traditional methods. Noteworthy papers include: The Model Context Protocol-based Internet of Experts framework, which enables LLMs to inherit structured wireless network management capabilities. The Shadow Wireless Intelligence approach, which integrates LLM-driven reasoning with retrieval-augmented generation to enable intelligent decision-making in covert wireless systems. The On-Device LLM for Context-Aware Wi-Fi Roaming, which uses high-level reasoning in the application layer to issue real-time actions executed in the PHY/MAC stack.